Where I think AI goes next and where the value lands: essays with a stated thesis, ecosystem maps, seed-stage deal memos, and forecasts that say in advance what would prove them wrong. Jump to any section below.
Essays
What comes next in AI, and where the value goes
Long-form arguments with a stated thesis, the evidence for it, and the signals that would prove it wrong.
Biocomputing and organoid intelligenceRead it bottom to top, and note how thin each layer is: this is a research field with a handful of companies, not a market yet.
Research consortia and fundersMost of the money here is still grants.
Public fundersGrants for organoid computing and early commercialisationNSF (BEGIN OI)NIHCalifornia Institute for Regenerative MedicineAustralian Government (Industry Growth Program)
Academic groupsField-defining papers and NSF-funded projectsJohns Hopkins (Hartung lab)UC Santa Cruz (Braingeneers)Indiana University (Guo lab)MITUniversity of Michigan
Private and philanthropic capitalFew disclosed backers so farHorizons VenturesSchmidt Futures
Software and trainingProgramming living cells is mostly stimulation, recording and analysis today.
Remote access and APIsOnly two commercial offers I could verifyCortical Labs (Cortical Cloud)FinalSpark (Neuroplatform)
Recording and analysis softwareAcquire and analyse neural activity dataMaxWell Biosystems (MaxLab Live)3Brain (BrainWave)Multi Channel Systems
Interfacing and electrodesBorrowed from neuroscience and drug-safety labs.
High-density microelectrode arraysThousands of electrodes under one cultureMaxWell Biosystems3BrainMulti Channel Systems
Multi-well MEA platformsMany cultures recorded in parallelAxion BioSystemsMaxWell Biosystems3Brain
Neuron-on-chip hardwareThe companies actually selling or renting biocomputers.
Integrated biocomputersLiving neurons on electrodes with life supportCortical Labs (CL1)FinalSparkBiological Black Box (Bionode)
Neuron-based sensingEngineered neurons as chemical detectorsKoniku
Wet-lab platformsCells and organoids, the raw material.
Human neuron and organoid supplyiPSC-derived neurons and cerebral organoid kitsbit.bioFUJIFILM Cellular DynamicsSTEMCELL Technologies
Organoid research labsGrowing and training brain organoidsJohns Hopkins (Hartung lab)UC Santa Cruz (Braingeneers)Indiana University (Guo lab)
Seed-stage deal memos, one for each of my big bets: the round, the team, what I would have to believe and the questions I would ask first. They live on the Venture page.
Each trend with how much momentum it has today, the evidence behind it and the signal I would watch.
Trend
Momentum
Evidence
What to watch
Inference keeps getting cheaper
Rising
Stanford's 2025 AI Index found the cost of running a GPT-3.5-level system fell more than 280-fold between November 2022 and October 2024. Stanford HAI, 2025 AI Index
Whether falling token prices widen application margins or get competed away, and which apps only work at today's prices.
Pricing that shifts from seats to completed work, and which incumbents lose seat revenue as a result.
Vertical AI in regulated industries
Rising
Harvey says 80% of Am Law 100 firms use it, and Abridge expects to support more than 300 large US health systems in 2026. Harvey, Sept 2026
Incumbent platforms shipping native versions, as Epic did with AI Charting, and whether startups move into revenue work first.
Robot foundation models
Early
Skild AI raised nearly $1.4B at a valuation above $14B in January 2026 to build one model for any robot body. Dow Jones via MarketScreener
The first paid deployments outside demos, and whether value pools in the model or in the robot maker.
On-device AI
Early
Apple opened its roughly 3 billion parameter on-device model to developers in 2025, with inference that works offline at no cost to the app. Apple Newsroom, Sept 2025
Which app categories move work off the cloud, and what that does to per-token revenue for model providers.
Energy and compute constraints
Rising
The IEA projects data centre electricity use more than doubling from 415 TWh in 2024 to around 945 TWh by 2030. IEA, Energy and AI
Power access as the gating factor for new capacity, and startups selling efficiency rather than more compute.
Stablecoins as payment rails
Rising
The GENIUS Act became US law on July 18, 2025, and Coinbase reported a record $20B average of USDC held in its products in Q2 2026. White House: S. 1582 signed
Stablecoin use in real payments and treasury flows rather than trading, and how rate cuts change issuer economics.
Biocomputing research
Uncertain
NSF put $14M into seven organoid intelligence projects in 2024, and the field still has only a handful of companies. NSF, Aug 2024
Peer-reviewed benchmarks of useful work per watt, and pharma buying these systems as research tools.
Nothing here is investment advice: these are my own notes on public information, written to sharpen judgment, not to recommend buying or selling anything.
Research
Forecasts and working papers
Each piece states in advance what would prove it wrong: calibrated probabilities, pre-registered hypotheses, or an explicit list of what the method cannot see.
Jun 2026
The Chokepoint Decade
Forecasting the collision of AI and modern mercantilism, 2026 to 2036
AI and mercantilism as one competition for five scarce inputs: capital, energy, critical minerals, semiconductors and skilled labor. 24 binary forecasts with objective resolution criteria, calibrated from 0.40 to 0.88.
Bridgewater Associates × Global Citizen, Forecasting the Future Challenge
2026
Measuring Buyer Intent from Costly Public Signals
Working paper
Weights every public buyer action by what it cost the actor to produce. Intent in log-odds with Good's weight of evidence, a hazard model so silence counts, and an explicit contract for what the instrument cannot see.
Working paper
Jul 2026
Predicting Fan Engagement Under Sparse Signals
A pre-registered cold-start design for the 2026 FIFA World Cup
Extends latent-attrition models into the cold-start case with a graph-structured prior. Seven falsifiable hypotheses registered before any data was fit.
Working paper and reproducible pipeline
Jul 2025
Forecasting the AI HealthTech Decade
Power, policy and wellbeing, 2025 to 2035
Health data and clinical AI as sovereign infrastructure. 20 calibrated forecasts and a macro feedback model tested against five country cases.
Independent research paper
Abstract
The ROI of Superfans
Lifetime fan value from social graphs and engagement heatmaps
1.2 million fan profiles and 150 million touchpoints, graph embeddings and survival regression, 88.4% out-of-sample accuracy against an RFM baseline.
MIT Sloan Sports Analytics Conference, past first-round review
Published on the products I build
Research reports, posts, guides and a manifesto I wrote for Searchalong and Keystone. Each opens on the product's own site.
A product idea, a team that needs another builder, a competition that needs a judge. Tell me what it is and I will say straight away whether I am the right person.